Oculus CES 2014 prototype "Crystal Cove"

novalkarnovalkar Posts: 6
edited January 2014 in Oculus Rift
This is a little resume about the information that was disclosed through news about the latest Oculus prototype that was presented to the press at CES 2014 during Day 1. I painted in blue the statements that are based in my personal opinion.

First of all, "Crystal Cove" is only a prototype and by no means is their final implementation nor DK2.

Here is a video of the kit being tested by a reporter from IGN: http://youtu.be/2kdcEwias1Y

As far as i've seen, there are 2 mayor enhancements that are being commented on the interwebs.

* Reporters who tested it say that it fixes the "motion blur", makes text more readable and it didn't produce nausea.
* The theoretical concept is explained here: http://blogs.valvesoftware.com/abrash/w ... in-judder/
* Is also explained in more simple words by Palmer in this video: http://youtu.be/AMPFXYGoNHI?t=30s
* They are using a special OLED screen to achieve this effect, Palmer stated that this was possible because the screen they used was able to work this way and that "It is not something that any OLED panel can do". [Source: http://www.polygon.com/2014/1/7/5284258 ... mysterious]
* OculusVR haven't officialy disclosed the screen provider or resolution, but most reporters say that it must be 1080p.

Current updated state of positional tracking solution
* The current prototype has several LEDs in the surface of the Rift and uses a camera to track them.
* The concept looks a lot like the one from this old video http://youtu.be/Jd3-eiid-Uw
* Since OculusVR haven't official disclosed more information about the full prototype tracking solution, my safest guess is that they are using a mixed setup, where they are still tracking the orientation of the headset and complementing that information with the one that is seen using the camera (like how the wii does with it's controllers).
* The reporters that tested stated that "head positional tracking" helped a lot in the immersion aspect.
* One of the main worries that the people have is that since the tracking is "vision based", they think that it might not work well in 360° (like in an OMNIdirectional threadmill *wink* *wink*)
** People tend to forget that this is just a prototype and that things like that can be solved using more cameras, more LEDs or using a completely different tracking solution
** Nothing is set in stone yet.
* I would love to see how this "positional tracking" solution compares to the one showed by Sixense in it's STEM Kickstarter page: http://youtu.be/UxSZ0fk_ubA

What are your opinions about the new prototype? did i miss something?


  • SuprM4nSuprM4n Posts: 81
    I believe the most cost efficient manner of "fixing" the vision based issue would be adding LED's to the bands that keep the Rift on your face. Then, the only way that the camera would be unable to see them is if you were facing the opposite direction and had your head all the way down looking at your feet, whilst also lowering your head... (Seems unlikely to me, and if you are doing that then I doubt it's helping with anything in game).
    Props go to Plixity for the sig.
  • An alternative approach is to get a second IR camera from Oculus and for Omni to include two camera inputs and a switch which selects which one to use. A 'which-camera' output (maybe two IR leds on the Omni frame) would also need to be passed along with the camera output to tell Omni that the orientation had changed 180 degrees.
Sign In or Register to comment.